Modelos predictor de la morosidad con variables macroeconómicas

Abstract

This paper presents a proposal that analyzes and anticipates the risk of the non-compliance of delayed loans using Arima models that help to identify the macroeconomic indicators associated with the outstanding debt in each segment of the Ecuadorian credit. These warnings will undoubtedly contribute to the construction of systems capable of anticipating payment defaults in advance. The analysis period contains a cycle of 2010-2015 with monthly values. The sample includes macroeconomic variables of each segment and the fundamental risks of the financial system. Five of nine generated models were validated with at least twelve months of anticipation in the study period.

Keywords

• Arima models
• delinquency rate
• early warning
• risk of default
• time series